Embed Size (px)
Citation preview
Lang. Soc. 10, 21-42. Printed in the United States of America
Idiomaticity and functional variation: A case study ofinternational amateur radio talk
DAFYDD GIBBON
University of Gottingen
I . THE DOMAIN AND METHODOLOGY OF THIS STUDY
The linguistic domain of idiomaticity poses many problems for the study oflanguage form, use, and variation. With selected aspects of idiomaticity as astarting point, I will attempt in this paper to develop a description of the use ofidioms as a segment of a more general theory of language use. Evidence for thisapproach is drawn from international amateur radio talk (IART) in English.
Before starting on the description of IART, I will outline the main features ofidiomaticity and a method for explaining contextual factors in language varia-tion.
/ . / . Idiomaticity
Criteria for idiomaticity are found in all areas of semiotics: language structure(phonology and syntax), meaning (semantics), and use (pragmatics); each areainvolves systematic but quasi-lexical enumeration rather than description that canbe generalized by rule. This paper treats relations between structures andfunctions in idiomaticity; semantic considerations are omitted (but cf. Gibbon, inpress). Terms like "stereotypy," "collocational stability," "frozenness," and"fossilization," are often used for the syntactic aspects and will be used herealong with the term "idiom" (cf. Makkai 1972).
Hockett (1958: 309) proposed a hierarchy of idioms, from the morpheme tolevels where "idioms merge imperceptibly into the sorts of discoursewhich . . . can reasonably be called literature.'' The thesis of the present paper isthat his idea can be generalized from literature to speech, and that "restrictedlanguages" (Firth 1968: 87) may involve idioms right up to the higher discourselevels.
The idiomaticity literature cannot be reviewed here, but a synthesis of relevantaspects can be given, grouped into subsentential and suprasentential levels. 1.Subsentential: a. phonological - rhyme, alliteration, phonaesthetic and synaes-thetic units (Smith 1925; Bolinger 1950); b. morphological (simple) - morphemesas minimal stereotypes (Hocket 1958; Katz & Postal 1963); c. morphological(complex) - nonproductive word compounds and composites (Marchand i960);d. phrasal - ill-formed, "unambiguous" idioms like by and large, to kingdomcome (Newmeyer 1974; Weinreich 1966); e. clausal - transformational anoma-lies, from "binomials" (bacon and eggs, eggs and bacon vs. kith and kin, *kinand kith; Malkiel 1959) to verbals (Gareth gave Gwyneth a piece of his mind,0047-4045/81/010021-22 $2.50 © 1981 Cambridge University Press
21
DAFYDD GIBBON
*Gareth gave a piece of his mind to Gwyneth; cf. Makkai 1972), along a scale oftransformational frozenness (Fraser 1970). 2. Suprasentential: a. simple dia-logue contributions ("speech acts") - proverbs, citations, interjections, excla-mations, tags, parentheses (Bublitz 1978; Sadock 1974; Searle 1976) and theprosodic stereotypes that mark some of these (Chao 1956; Gibbon 1976a; 274ff.,1976b; Ladd 1978); b. complex dialogue contributions - set monologues likestories, jokes, poems, often embedded in practical/aesthetic interaction types(selling, music, etc.); c. simple transactions ("exchanges") - phatic sequenceslike greetings, probably Ervin-Tripp's "routines" (1964: 89), and uptake loops("Which? - That. - Oh!," cf. section 3.4, below); d. complex transactions -fixed performances like plays, liturgies, often with practical/aesthetic interaction;Leech's "disjunctive and abbreviated modes" (1966: 90), the "restricted lan-guage" end of Crystal and Davy's "scale of utilization" (1969: 63). Suprasen-tential idioms have been called "pragmatic idioms" (Burger 1973; Giilich &Henke 1980).
It is essential to the present argument that the hierarchy be understood to coverspecific composite speech forms with a common property of structural frozen-ness (nonproductivity, membership of enumerable lexical sets) and notfunctionally defined dialogue strategy types or genres. Since, however, the "de-fining contexts" for specific idioms are very restricted (Hockett 1956: 223),reducing form-context relations to near-isomorphism, at higher levels stable col-locations may be seen as prototypes of speech act types, genres, registers, andso on.
Two further properties of these forms must be noted. First, idiomaticity isscalar, ranging from weakly to strongly idiomatic combinations (cf. Fraser'sfrozenness hierarchy). Second, idioms contain "leaky points," as in discontinu-ous idioms (allowing insertion of limited kinds of nonidiomatic material; cf.bring X to light) or in idioms where small amounts of free variation are tolerated(bury the hatchet ltomahawkl*axe). These properties also hold for higher levelstereotypes.
1.2. Constitutive systems and functional variation
Functional variation is the adaptation of utterances to the immediate context ofuse. Its domain is often referred to by the cover term "register" (Ellis & Ure1969; Leech 1966); adequate context descriptions, however, are not given bysituation lists (lecturing, telephoning, radio communication, etc.) or varietytaxonomies (lect, style, genre, register, etc.), even for stereotypic forms whose' 'defining contexts'' are restricted by comparison with the weaker ' 'determiningcontexts" for freer varieties. A need for a more basic frame is implicit indefinitions of taxonomic categories in terms of different "primitives" such as"participant," "user," "use," and in the hybrid taxonomies proposed byHymes (e.g., 1972). An explicit, coherent system of such primes is needed to
22
IDIOMATICITY AND FUNCTIONAL VARIATION
make such definitions useful: if a dialect is a "variety according to user" and aregister is a "variety according to use" (Halliday, Mclntosh, & Strevens 1964:77), where is the border between the language of boilermakers and the languageof boilermaking?
A basic set of constitutive factors for communication was proposed within thePrague school by Biihler (1934) and extended by Jakobson (i960); the approachhas also influenced Hymes's ethnographic method (1975). These sets were origi-nally used to define language functions, - speech appeal as hearer-orientation(Biihler 1934: 29), for example. Such sets may be arranged along a scale ofrelative idealization: Chomsky's "ideal speaker-listener" (1965: 3) is high on thescale; lower are Biihler's sign, speaker, hearer, context; then come Jakobson'sspeaker, hearer, contact, context, message, code, and the "indices" (sets ofcontext coordinates) in formal semantics (e.g., Lewis 1972). The sets are oftenhybrid in content, being based on unanalyzed intuitions about context compo-nents, rather than on a structured theory; factors such as hearer, participant arethemselves functions, though they are at a different level from those they defineand are in need of further explanation.
A more highly structured descriptive approach will be developed here in threestages.
First, structured sets of constitutive categories are developed by making suc-cessive distinctions within speech events:
Si: <SPEECH EVENT>S2: <UTTERANCE, CONTEXT>S3: <UTTERANCE, <PERSON, ENVIRONMENT»S4: <UTTERANCE, <PERSON, <CHANNEL, S E T T I N G > »S5: <UTTERANCE, <<INPUT, OUTPUT>, <CHANNEL, S E T T I N G > »
The categories may be analyzed further as the empirical domain of discourserequires; utterances, for instance, are complex gestural wholes to be analyzedinto kinesic, paralinguistic, prosodic, and segmental components.
Second, these category sets are developed into constitutive systems by assign-ing to the members of categories properties and relations from a set of contextualparameters. These parameters may be roughly classed as natural (e.g., physical,perceptual) or conventional (e.g., codified or uncodified group consensus oninterpretation); a related distinction is drawn by Searle (1969: 5off.) between"brute" and "institutional" facts (cf. also Gibbon 1976a: 28). The followingexamples illustrate parameter values for the idealization level S4. A PERSONenters into natural (perceiver), conventional relations (addressee, third-party-listener, eavesdropper) or hybrid relations (receiver, decoder), all implicit in the
23
DAFYDD GIBBON
CHANNEL•
SETTING
, *
INPUT OUTPUT
J
\
, vINPUT
f
OUTPUT
>f
PERSON
ENVIRONMENT
Figure i: Sketch of dynamic relations between components of a functional constitutive system.
standard terms "hearer," "listener." A SETTING involves natural or conven-tional objects (cf. "cave" vs. "home"). A CHANNEL has natural (visual vs.oral, noisy vs. clear, long-range or teleglossic vs. short-range), conventional("It might be more suitable to tell her in person"), or ambivalent (private vs.public) properties.
Third, in order to explain functional variation, dynamic relations within aconstitutive system can be defined as a controlled, self-regulating process ofcontinuous adaptation to and modification of the ENVIRONMENT by PER-SONS, as shown in Figure i. Although such configurations have often beensuggested for the description of language use (from Shannon 1949, throughHockett 1955, to recent psycholinguistic studies) they have often lacked detailedempirical and theoretical motivation. A dynamically interpreted constitutive sys-tem may be termed a "functional constitutive system," and the "functionalcycle'' is its operative principle of continuous adaptation to and modification ofthe environment. The functional cycle forms an empirically motivated frame forexplicating dynamic notions like functional variation, or types of variation ininteraction such as "code-switching" (Blom & Gumperz 1972). It provides asuitable basis for describing notions listed in situation lists and varietytaxonomies, which are simply cover notions for configurations of somefunctional constitutive system. They have no primary empirical status, thoughthey possess considerable heuristic value and intuitive appeal (Hymes 1972).
2. INTERNATIONAL AMATEUR RADIO TALK: THE CONTEXT
National and international laws on radio communication, as well as publicationsof national amateur radio organizations such as the American Radio RelayLeague and the Radio Society of Great Britain, describe the essential contextualfeatures, give lists of low-level codified phrase stereotypes and often containfragmentary notes on I ART usage. These notes are largely prescriptive, basedpartly on law, partly on etiquette. Typical topics for metalinguistic comments,
24
IDIOMATICITY AND FUNCTIONAL VARIATION
also heard on the air, are the justification for using telegraphy abbreviations intelephony, or the desirability of CB-type slang (Smith 1979). These topics relateto PERSON-oriented jargon in the technical sense (linguistic markers with phaticfunction for a restricted group); jargon is, of course, a common topic in other folklinguistic domains.
The present account is descriptive, based on lengthy experience of IART andanalytic observation over a five-year period; the results have been used to planEnglish classes for radio amateurs in Germany. To hear IART in use, a short-wave receiver with beat-frequency-oscillator (BFO) is required; the relevantfrequency bands are 20 meters (14,000-14,350 kHz) and 15 meters (21,000-21,450 kHz); exact frequency allocations may vary from country to country.Further information may be found in the official and semiofficial sources notedabove.
2.1. Summary of the main contextual factors
The main contextual factors for the kind of IART analyzed here are summarizedin the following taxonomy of situation dimensions. 1. PERSON: a. contacts -nonce-communication between strangers who do not seriously expect to meetagain (vs. "skeds," prearranged schedules for friends or network traffic); b.institutions - constraints by international agreements and national regulations onCHANNEL properties (transmission modes, frequencies, power), PERSONproperties (see Section 2.2) and UTTERANCE semantics (CHANNEL-orientedtopics or personal small talk, no commercial transactions or propaganda,frequent self-identification). 2. CHANNEL: technically aided vs. unaided,and so forth (cf. Section 2.3). 3. UTTERANCE-ENVIRONMENT relations:CHANNEL-oriented metacommunication for equipment testing, studying prop-agation conditions, "collecting" contacts (vs. rag-chewing with old friends,contests, emergency traffic, etc.).
It will be shown below how these factors constitute a powerful set of naturaland conventional constraints on IART speech; an appropriate term for a highlyconstrained variety like IART is "restrictive register." Both natural constraintson establishing and maintaining contact under difficult physical conditions andconventional constraints from the codified regulative institutional framework oruncodified constraints on contacts between strangers contribute in different waysto stereotypy of UTTERANCE forms at all levels, even giving whole transac-tions a collocationally stable, "ready-made" and therefore essentially lexicalcharacter, with "leaky points" (cf. Section 1.1) where discussion of technicalproblems or small talk occurs.
A peculiarity of IART is that ENVIRONMENT reduces to the categoryCHANNEL, as the summary above shows, with no other SETTING features.This is an oversimplification in view of the leaky points, but a justifiablegeneralization over the characteristic "clear cases" of IART (cf. Ervin-Tripp's
25
DAFYDD GIBBON
"relatively pure code," 1964: 90). This simplifies the relevant series of constitu-tive category sets:
S4': <UTTERANCE, <PERSON, CHANNEL»S5': <UTTERANCE, «INPUT, OUTPUT>, CHANNEL»
The following subsections treat the categories of PERSON and CHANNEL inmore detail.
2.2. Properties and relations for the category PERSON
The values of PERSON parameters in English language IART will be sketchedinformally. The community of IART speakers is (apart from a shared personalinterest in the hobby) characterized institutionally: the members are adults (usu-ally over eighteen) who are licensed by examination on technical and legalmatters and on operating technique; this is the defining feature of the radioamateur in international law. No definition by nationality or class is possible,though a bias toward the industrialized nations and elsewhere toward the coloniz-ing classes exists. More local social attributes derive from the status oftechnological hobbies in different societies; the most obvious is male predomi-nance. Otherwise the range of social groupings is unrestricted and includeslinguists, housewives, and heads of state as well as the "typical" male artisan orclerical radio ham. The community is multinational, numbering over a millionworldwide; just over half are non-native speakers of English, many havinglearned English via amateur radio. Contextual constraints on IART in otherlanguages are identical to those on English, and the linguistic effects appear to besimilar.
Communication in IART is generally dyadic, though larger groups may arisespontaneously or by prior arrangement. An interesting feature is that any radioamateur is equally entitled to a. initiate a transaction (subject to common senseand politeness conventions, e.g., about when a frequency counts as "oc-cupied"); b. participate in a transaction initiated by a general call, or a restrictedcall if the radio amateur falls into the specified category; c. alter the initiativefunction as required by the communicative process. Participant equality, rare inother restrictive registers (cf. Section 4), is due to the near absence of statuscriteria; two examples of higher status are network control stations and rare dx(long-distance contacts).
2.3. Properties and relations for the category CHANNEL
The CHANNEL used in IART has the following properties. 1. Natural: teleglos-sic (vs. face-to-face); simplex (nonsimultaneous transmission and reception vs.duplex, etc.); unimodal (vocal vs. multimodal vocal + visual, etc.); technically
26
IDIOMATICITY AND FUNCTIONAL VARIATION
PERSON
CHANNEL/
OUTPUT
(manual) (oral)
INPUT
(auditory)
tactilesubchannel
acousticsubchannel
1 KEY Y--\ MICROPHONE I A HEADPHONESY-ielectrical
subchannel
Y 1RECEIVER
electromagneticsubchannel
Figure 2: Channel structure in IAR talk and morse telegraphy.
aided (vs. unaided); transient (vs. storage) medium; electromagnetic (vs. acous-tic or material) propagation; shortwave (transcontinental vs. local, e.g., VHF/UHF propagation); direct (vs. relayed via balloon/satellite-mounted, or fixedrepeaters). 2. Conventional: legally restricted in mode (telephony, telegraphy,etc.), power, frequency; public by legal definition (unlike a "public" telephoneservice, which is public only in access or ownership). Natural and conventionalaspects of the public CHANNEL relate to constraints on IART usage. Oneconsequence is that participants may have an audience of unknown size andidentity; this often disturbs new operators and fosters partially codified topicconstraints, but it is a sine qua non for initiating new contacts and otherwise haslittle effect on average "clear case" contacts. Also ambivalent is CHANNELdirectionality: both (natural) directional radiation and (conventional) verbal spec-ification of target PERSON group (cf. Section 3.3.) may be used to restrict theset of addressees.
Natural aspects of the functional cycle in relation to a single participant areillustrated in Figure 2. CHANNEL structure is complex, each substretch con-tributing its own properties.
The CHANNEL category is related to that of UTTERANCE in four mainareas: (1) as the topic of metacommunication (CHANNEL-as-SETTING); (2) inthe relation of simplex CHANNEL to turn-taking routines; (3) in the relation of
27
DAFYDD GIBBON
PERSON CHANNEL UTTERANCE
QSL-cardexchange
Uptake-securingtechniques &
idioms
Framingidioms
Topic-orientatedidioms:
CHANNEL, small-talk
J
Figure 3: Outline of major <UTTERANCE, CHANNEL> relations C/xC means metacommunica-tive semantic relations).
teleglossic CHANNEL to deictic patterns; and (4) in the use of redundancy-raising patterns to combat noise and low channel capacity. The majorUTTERANCE-CHANNEL relations are sketched in Figure 3. Most communica-tion is therefore CHANNEL-orientated metacommunication of some kind (ab-breviated "jtiC" in the figure)-that is, phatic in the narrow sense of Jakobson(i960). A higher level of metacommunication is present in written documenta-tion of contacts (log-book entries and exchange of QSL-cards to confirm contactsfor various technical and conventional purposes).
Like PERSON-CHANNEL relations, the UTTERANCE-CHANNEL relationscan create problems of "user noise" (fragmentary knowledge of or low fluencyin I ART) for newcomers to the variety.
3 . LINGUISTIC PROPERTIES OF IART
In this section a sample of IART data will be presented, with an informalcommentary on the relevant properties; subsequently the lexicon and structure ofIART will be described more systematically and then summarized in terms of thefunctional cycle described in Section 1.2.
3.1. Selected fragments of IART contacts
The following excerpts from tape transcriptions illustrate dialogue patterning(initiation, turn control, uptake control, termination) and CHANNEL-as-SETTING utterance semantics (i.e., "topic") in IART. Some straightforward
28
IDIOMATICITY AND FUNCTIONAL VARIATION
transcription conventions are used: hyphens indicate spelling out; participants arelabeled Pi and P2 in order of appearance; to preserve anonymity, the individuat-ing suffix letters in the call signs are replaced by xx, or xray xray if aheterophonic spelling alphabet is used (perhaps an unnecessary precaution inview of the public CHANNEL property of IART).
1. Pi: . . . Ontario henry 2 xray xray calling cq dx, Ontario henry 2 xray xrayis by for a call.
P2: Ocean henry 2 xray xray could I have a quick report old man, echoindia 9 xray xray.
Pi: Echo india 9 xray xray, Ontario hotel 2 xray xray, you're five nine plus20 to 30 db, fine - five nine plus 20 db in the vicinity of Helsinki, ei9xxoh2xx.
In this initial fragment of an IART contact, P i , in Finland, makes a generalcall (cq) directed to potential long-range partners (dx); by (standing by) is anexplicit turn-yielding signal. P2 (Irish Republic) is not in the addressee group,being in Europe, so his reply is functionally "marked"; this status is formallyreflected in quick and the abbreviated turn-start and turn-end. Standard turn-starts and ends are present in P2 's second contribution: addressee and speakercall signs in that order. The numbers denote values on metacommunicativejudgment/measurement scales of received signal properties: readability (values1-5), strength (1-9).
2. Pi: . . . your signal here is five by nine, five by nine, very fine signal fromEngland, very fine signal. The name here is Ken, kilowatt echonovember. We're located in Taylor, Missouri, t-a-y-1-o-r Taylor Mis-souri, and er let's see em er back to you, I didn't get your name, golf 3xray xray, kaxx.
This medial turn shows redundancy-raising strategies: repetitions, alphabeticand heterophonic spelling out, and the initial segment of an uptake loop (seesection 3.4). Two deictic conventions are also shown: first person plural forspeaker, relatively common, and generalization of proximal local deixis to firstperson, ascribable to speaker-addressee distance in the teleglossic channel - thename here for my name.
3. Pi: . . . and Mario please correct your qth, is that Ontario radio indianovember delta delta alpha, is that roger?
P2: Negative, negative, Alex [unreadable] is this okay now, go ahead.Pi: er olinda, ocean london india november delta alpha, is that roger?P2: Okay okay one hundred percent copy now. Seventy-three to you good
luck to you and the family and I'll see t- I hope to talk to you againsome [unreadable].
29
DAFYDD GIBBON
In this exchange, Pi (USSR) initiates an uptake loop to elicit his partner'slocation (qth) in Brazil. On successful completion of the loop, P2 initiates atransaction terminating routine: seventy-three is a relict of older telegram codes(best wishes); the stereotypic nature of good luck to you and the family is obviousin the context of nonce communication. The same goes for the frequently heardhope to talk to you again; the chances of this happening are remote unless a"sked" is arranged.
4. Pi: Lima zulu 2 xray xray from victor kilo 3 xray xray returning. RogerMario, okay on your qth okay on your qth, papa lima echo victoria echonovember, roger and thank you very much indeed for the contact. Youare now strength 6 Mario, you are now 5 strength 6. All the very bestfrom Australia. London zulu 2 xray xray from victor kilo 3 xray xraynow clear. Cheerio Mario.
P2: Yeah okay Bill, thank you very much again. All the very best to youand yours and to your wonderful [unreadable] best of seventy-three andhave a good weekend, Bill. vk3 xray xray, Iz2 xray xray signing.Bye-bye.
After uptake confirmation (roger) on the location of P2 (Bulgaria), Pi (Aus-tralia) initiates the termination routine; the features resemble those in the previ-ous example. A standard phatic strategy to reduce distance between strangers inIART is use of first names; frequent repetition of these, as here, supports thisstrategy. Discourse relevance of a morphological feature is shown in spellingstrategies for call signs: the general regional prefix (e.g., vk2, Iz2) is oftenspelled out in the briefer alphabetic form, whereas the person-identifying suf-fixes (here uniformly xx) are spelled out more prominently with a heterophonicspelling alphabet. The principle is similar to that of stress assignment to thespecifier constituent in many compound words. At the clausal level, a similarforegrounding strategy is used, with brief alphabetic spelling for the speaker'scall sign and the fuller version for the addressee's call sign. The common de-nominator for these prominence-producing strategies is the given-new, or topic-comment, distinction (cf., e.g., Halliday 1967), linked with a semantic relation(hyponymy) in the first case and with pragmatic conventions (politeness) in thesecond.
5. Pi: . . . okay on your coffee. Well we haven't yet er got around to ourevening cup o' coffee er I say "we" - my wife's been sitting along-side me here [introduces wife].
P2: Okay vk3 xray xray, sm6 xray xray. Fine business copy, Don, and helloto Carolyn in the background there. You certainly have found my qth onthe map there, Don . . . well, er, fine on the rig, Don, doing a very nicejob for you, no trouble whatsoever. Seventy-threes to you and seventy-threes and eighty-eights to Carolyn in the background there and hope to
30
IDIOMATICITY AND FUNCTIONAL VARIATION
have another chat with you, perhaps a longer one next time, Don.xray xray, sm6xx is clear and I will qsl you one hundred percent, erDon. Bye-bye.
Pi: er Cheerio [unreadable] and thanks very much for the qso. sm6xxvk3xx off and clear and er qrz europe, qrz [noise from several newcallers].
Apart from several features already noted (uptake confirmation idioms okayon ... , fine on ... ; first person plural, this time a literal use, functionallymarked in IART and formally marked here by a metalinguistic comment;generalization of local to personal deixis, this time distal: there = you/your, etc.;termination stereotypes) example 5 shows a leaky point near the end of thetransaction with small talk about the immediate environment. The non-CHANNEL SETTING is non-null (Pi's wife, coffee, map). Eighty-eights is arelict of telegram code (love and kisses), used as a polite stereotypic greeting toor from females, without literal connotations. The termination routine refers toQSL-card exchange; qsl has the general meaning of uptake confirmation (cf.roger), but is frequently used in this specialized sense. Stereotypic third personlistener reference is made by . . . in the background there, twice. Pi 's finalcontribution immediately initiates a new transaction with the question qrz? (whois calling me?), sometimes used as a substitute for cq.
The main points to note are the codified or otherwise stereotyped forms atvarious linguistic levels, particularly adherence by participants to a core transac-tion that may be expanded in restricted directions at certain fixed leaky points.
3.2. The lexicon
IART vocabulary is restricted both semantically - covering mainly CHANNELproperties and CHANNEL-PERSON relations, actions, and interactions - andmorphologically, in its numerous abbreviations and other nonstandard wordstructures; a further restriction is syntactic: items are often ambivalent betweenclausal and subclausal categorial status. The morphological oddities were origi-nally introduced to compensate for low channel capacity in Morse telegraphy: the2-bit (=£4 item) code alphabet requires longer syntagmatic groups to encode the6-bit (=£64 item) Roman alphabet, numerals, and punctuation, as well as beingslower in production and perception (in amateur traffic, generally 60-120 lettersper minute). In telephony, they have two secondary effects: 1. as group-identifying jargon (cf. section 2); 2. language neutrality, due to the arbitrarinessor at least nontransparency of most items.
Five types of characteristic lexical items can be identified: (1) Q-signals, (2)abbreviatory codes, (3) whimsical abbreviations, (4) numerical codes, and (5)alphanumeric scalar codes.
First, the Q-signals, perhaps the most distinctive, are also used in officialtelegraphy and telephony and in private radio (CB). Their general structure is
31
DAFYDD GIBBON
Qxx, with xx standing for some biliteral sequence; in telegraphy, they are "prop-ositional act" idioms (Searle 1969) with choice of illocutionary force in requests(for information or action, marked by a telegraphic question mark in the firstcase) or statements (cf. A.R.R.L. 1977: 647):
QRP: Shall il decrease power? Decrease power.QRZ: Who is calling me? You are being called b y . . . (on . . . kHz).QTH: What is your location? My location is . . .
The dots show expansion points for inserting restricted types of further informa-tion. Syntactic aspects of Q-signal sequences are treated in section 3.4. Intelephony, relative lack of terseness has induced a functional shift from thephrase-structure category of Sentence idiom to lower ranks like Noun, Adjective,Adverb:
QTH: [smy location is . . . ]—» [\my location]
QSL: [s/ am acknowledging receipt] -* / ^<wri"en> comfirmation]( [yconfirm (in writing)]
QRP: ^Decrease power] -* { [A^W power {-ed)]\ LAtivW"1 low power]
Examples of functionally shifted uses are: my qth is London, I will qsl via thebureau, I am using a qrp rig, I am working qrp.
Abbreviatory codes, the second type, derive from English or French words.One subtype retains the first letter of an English word, replacing the rest by adummy x, as in tx, rx, dx, and wx for transmitter, receiver, distance, andweather. Others are acronyms, like ut {universal [Greenwich mean] time), withthe arbitrary synonym z (spoken as zulu, e.g., 10:45 zulu), or abbreviationspreserving syllable information, such as cfm and abt for confirm and about.
The third type uses homophony between alphabetic letter names and words,such as cq {seek you), xyl {ex-young-lady, i.e., wife), or (only in telegraphy)bcnu {be seeing you) and cuagn {see you again). These are reminiscent ofchildren's spelling riddles like 2YYUR2YYUBICUR2YY4ME. There is even anexplicit idiom for laughter in telegraphy: hi, a rhythmical code sequence of fourplus two dots; it has been adopted by some operators in telephony with the lettersspelled out /eicai/, or /hai/, or reduplicated as /haihai/. The original telegraphicuse, as an addressee-initiated uptake confirmation (see section 3.4) with addi-tional positive evaluation, has changed in telephony to a speaker-initiated uptake-eliciting signal used as a tag, like isn't it, nicht wahr, or n'est-ce pas, withevaluative function.
The fourth type, numerical codes, is a relic of older wire telegraphy. Theexamples seventy-three and eighty-eight occurred in section 3.1. Interlanguageinterference may be observed with the German whimsical number code 55 (in
32
IDIOMATICITY AND FUNCTIONAL VARIATION
telegraphy, five plus five dots), sometimes transferred by German hams to En-glish, puzzling their non-German speaking interlocutors. It is based on an au-tological pun: a series of ten dots is viele Punkte, that is, many dots; Punktmeans both dot and point (in a contest, etc.), giving the derived meaning manypoints, that is, every success, all the best, (in a contest, etc.), which has in turnbecome generalized to other situations.
The fifth type, alphanumeric scalar codes, was also represented in the datafragments; the telegraphic expression might be ur sigs rst 599, for instance,meaning your signals are value 5 on the readability scale, 9 on the strengthscale, 9 on the tonal quality scale. Exact details are given in the sources alreadymentioned.
Some letter codes were originally arbitrary, based on perceptually distinctiveMorse rhythms and later received acronymic folk etymological interpretations.The best known example (prohibited in amateur radio and reserved for officialemergency services) is the originally unanalyzed code sequence • • • • •, laterconstrued as sos and then as Save Our Souls. It is possible that ok (okay) has asimilar origin in nineteenth-century telegraphy (codified as agreed; cf.H.M.S.O. 1938: 91), though its origins are much debated.
Besides the abbreviations, the vocabulary includes standard alphabetic andcodified heterophonic letter names for spelling out, ordinal and cardinal num-bers, and stereotypic expressions at the various levels illustrated in section 3.1.Common expansions of the core transaction require expressions for frequenciesand other physical measurement scales and units, components of radio equipmentand their functions, time, date, weather, and so on. Further vocabulary is intro-duced at expansion points as required.
Some idiosyncracies of "closed set" grammatical formatives were noted insection 3.1, the most obvious being deictic terms. Another item to be noted is the-ing suffix as in . . . listeningl(standing) by /returning/calling, used together witha call sign as explicit turn-status markers.
3.3. Contribution syntax
The word and sentence levels are similarly lexical in character, allowing, forinstance, the simple Q-signal functional shift described in section 3.2. Contribu-tions are composed of such elements in stereotypic ways that may initially besummarized in phrase structure schemata. The contexts in which contributionsoccur have two main dimensions, which are symbolized here as feature contexts:(1) the PERSON involved, indexed for identity as [/ PERSON] (cf. the Pi andP2 of section 3.1), with [1 PERSON] being reserved for the transaction initi-ator; (2) status in the transaction context, initial or noninitial exchange, and soon, as [± INITIAL]. A third special aspect is category iteration, shown by astandard rule schema notation. These notational devices are provisional short-hand for more complex structures outlined in 3.4. This hybrid PSG contains the
33
DAFYDD GIBBON
following rules (the category names are mnemonically self-explanatory; cf. alsothe lexicon below):
1. CONTRIBUTION -> <START ((INFO)")'"> END / < [ - INITIAL]>2. START -» (IDENT)" (CONFIRM), n>03. END -> (CONFIRM?) (IDENT)'' CLOSURE, n>a4. IDENT -» (CALL,)" LINK (CALL;)'" (ROLE;) / \j PERSON], i?j,
n, m>0SPECIFIER/ a t - c ^ i r i c K \ \ / r ,
[GENERAL / M ; ~ ~ : 7 | , [_LINK .. . ] ]\ T A T I Cir .M / L ~
PERSON]
5. C A L L - I ™ " U C A L L - S I G N J / ' L + INITIAL J[CALL-SIGN J
The two levels of bracketing round INFO allow CHANNEL-oriented itemiteration (inner) and iteration of the whole category (outer) to accommodatedifferent information types and possible expansion points. Transaction-initialcontributions are treated, somewhat arbitrarily, as if they contain only an ENDconstituent (cf. section 3.1, item 1). Stylistic or "marked" features noted in thedata are omitted from the PSG, as are uptake loops.
A PSG is possible only as an <UTTERANCE, CONTEXT> categorial hy-brid, under the simplifications just noted, and because of the near-isomorphicform context relations of idioms in highly restrictive registers like IART. It doesnot work without these assumptions or for other registers. Items like START,END, and their constituents are intruders from the higher transaction level andwill receive more attention in section 3.4. A PSG provides a useful notation forsummarizing contribution from a limited perspective, however. The following isan illustrative lexicon fragment (cf. section 3.4, Fig. 5, for a specimen struc-tural diagram):
LINK: 0 , from, this is...
GENERAL: cq, qrz, qst (general announcement to all listeners);SPECIFIER: dx, Africa, Pacific, FloridaCALL SIGN: country prefix: G3-, K i - , . . . , 5X5-, 9G1-, etc.;
individuating suffix: - A , . . . , -777..CONFIRM: roger, okay (in first exchange, greetings: good morning, etc.)CONFIRM?: how do you copy?, etc.;CLOSURE: over, go (ahead);ROLE: calling, (standing) by, listening, returning;INFO: a. <UTTERANCE, <PERSON, C H A N N E L » relations: you are
five and nine, etc. (cf. sections 3.1, 3.2);b. <PERSON> properties, e.g., first name;c. <PERSON, CHANNEL> relations, e.g., qth (location).
34
IDIOMATICITY AND FUNCTIONAL VARIATION
3.4. An approach to the description of transaction structure
Transactions involve two or more PERSONS, requiring more complex descrip-tions than monosystemic PSGs can offer. Their main properties in IART, beyondthe contribution level, are explicit metacommunicative framing and uptake-securing. Contribution-level metacommunication is CHANNEL-as-SETTINGsemantics; framing and uptake-securing pertain to the pragmatics of CHANNELuse.
Framing, comparable to transaction-level "configurative functions" ascribedto intonation by the Prague school (e.g., Danes i960), is the marking of transac-tion component boundaries relevant to turn-taking, for instance, by IDENTifica-tion patterns, or participles in LINKS between these. Framing idioms includeover, go (ahead) (contribution end), (signing) ((off and) clear), (transactionend). The unimodal simplex CHANNEL prevents simultaneous addressee feed-back and addressee-originated framing like turn-bidding, which, with noise con-siderations, promotes explicitness in turn-yielding.
Uptake-securing (Austin 1962: nsff.) , that is, obtaining comprehension as aprerequisite for any speech act, is special semantics-based framing dependent on"normal input and output" preparatory conditions for speech acts (Searle 1969).IART speech can be explained functionally as being subject to "marked" or"abnormal" preparatory conditions based on <<INPUT, OUTPUT>, CHAN-NEL>relations (cf. section 2.1). Similar "metalocutionary" functions of pro-sodic devices may be observed, as in calls or other unaided teleglossic speech, inparalinguistic surrogates for these, or in ceremonial public speech, includingchants, and so forth (cf. Chao 1956; Gibbon 1976a; section 4.3; Ladd 1978).
The CHANNEL-orientation of uptake-securing can be further analysed as (1)OUTPUT-orientation: (a) linear (spelling-out, use of predictable explicitstereotypes, clear enunciation, occasional use of prosodic features of unaidedteleglossic speech (e.g., call contours), (b) cyclic (repetitions), (c) technical(filtering, power increase, speech compression) or (2) INPUT-orientation: (a)linear (roger, okay/fine (on . . . )), (b) cyclic (uptake loops to elicit repetition/expansion, etc., from previous speaker), (c) technical (tuning, filtering, low-noise amplification).
The OUTPUT-oriented patterns were already noted in section 3.3. INPUT-orientation requires functional constitutive systems (section 1.2) with more thanone PERSON and with a structure like Shannon's "correction channel" (1949:68), containing an additional "observer" to provide "correction data" for a"correction device." Uptake loops provide verbal feedback between two PER-SONS; they have been discussed under other names (cf. Hall's "adumbrations,"1964; Yngve's "back channel," 1970; the "repair mechanisms" of Sacks et al.,1974), but the transaction structures usually suggested are not adequate to ac-commodate them. More structure is needed than the "two turns at a time" binary
35
DAFYDD GIBBON
PERSON
C. Control oftransactiondevelopment
CHANNEL
(INPUT heuristics)
INPUT
METACOMMUNICATION
(uptake)
negative, i roger,etc. ! etc.
(framing)
over,etc.
INFO
«i + i
D. OUTPUT
i (REPLY)
Figure 4: Outline of functional constitutive system with construction of internal <INPUT, OUT-PUT> relations to explain uptake loops.
syntax of Sacks et al. or the ternary structures of Sinclair et al. (1975: 5Off.); thepurely turn-oriented approaches (e.g., Duncan 1973) stop short of integratingturn-taking with linguistic patterning.
The following schema summarizes uptake loop structure:... (<DISCON-FIRM REQUEST, REPLY>)\ CONFIRM... An uptake loop is a structuredquadruple of contributions: (1) a first exchange with a contribution containing anegative uptake signal (DISCONFIRM, e.g., negative, bad copy) and a RE-QUEST for repetition, followed by a REPLY; (2) a response superordinate to thewhole first exchange, containing a positive uptake signal (CONFIRM, e.g., qsl,roger, okay). The first exchange, the uptake loop proper, may be iterated as
36
IDIOMATICITY AND FUNCTIONAL VARIATION
required. Such loops occur freely after turn-switches, either as the next exchangeor initiated inside a longer contribution; distributionally, uptake loops may thusbe seen as transaction-level parallels to contribution-level parentheses. Thesuperordinate CONFIRM items may occur in any contribution outside a loop(partial confirmation, like / got the first letter, but... being a bridge to a newloop), an indication that uptake loop potential may be intrinsic to all nonloopcontributions, contingent on uptake evaluation by the addressee, but not alwaysrealized. On this assumption, the functional constitutive system of Figure 4 maybe constructed in order to account for the facts.
In this system, an utterance analysed as uj (by heuristics ignored here) entersthe system at A and is evaluated against current expectations derived from a previ-ous utterance M, at B. On positive evaluation, Uj passes to C for further develop-ment in a planning function that yields M1+I , which is transmitted via the syntax andphonology of D. The evaluation itself may trigger a metacommunicative signalaccompanying w(+, and referring to w, , or on its own as a pure feedback signal ofno w(+, is generated. On negative evaluation, INPUT-oriented uptake strategiesmay be triggered; if necessary, a loop is initiated and iterated (subject to furtherinterest/motivation/frustration factors). A loop initiation may also trigger OUTPUT-oriented strategies in the previous speaker.
The structure of this system accommodates the peculiarities of IART transac-tion structure; contextual expectations such as those formulated in the hybridPSG of section 3.3 are defined and revised in C as a function of time; compari-son with actual interpreted contexts at various levels (of which the uptake com-parator is only one) provides contextual expectations for continuing the transac-tion. Part of these expectations are the institutional and uncodified conventionson <UTTERANCE, CONTEXT> relations governing UTTERANCE semantics(e.g., topic), framing, and other aspects of CHANNEL use. The format of thispaper does not allow definition of an appropriate functional constitutive systemin full, but this outline, together with the systematic descriptions of the threemain levels of such a system (categories, relations, dynamic interpretation) inprevious sections, is sufficient to show its basic properties.
To illustrate the stereotypic core structure of transactions in this register, aspecimen qso (transaction, contact) has been constructed and is shown in Figure5. For reasons of space, a CHANNEL constraint on academic papers, the subva-riety of Morse telegraphy (in alphabetic transcription), IARMT, was used; al-though not all abbreviations used in the idioms have been documented above, theform-context isomorphism of the restrictive register will enable readers to dotheir own decoding. Parts A through D of Figure 4 and the PSG categories ofsection 3.3 yield relational "defining contexts" for the idioms. Particularlystriking is the regular "architecture" of transactions, with symmetrical ex-changes reflecting equal participant status (section 2). Use of the wider resourcesof English face-to-face communication would rapidly result in a conflict with therestrictive forces of the CHANNEL and the institutions that define the commu-
37
TE
LE
GR
AP
HY
IDIO
MS
TR
AN
SA
CT
ION
S
TR
UC
TU
RE
(WIT
HO
UT
LO
OP
S O
R E
XP
AN
SIO
NS
)C
ON
ST
ITU
TIV
E
SY
ST
EM
cq d
x d
e g4
fhz
-k g4
fhz
de W
4bc
-kn
IDE
NT
CL
OS
UR
EID
EN
TC
LO
SU
RE
de g
4fh
z ID
EN
Tgm
es
tnx
fer
call
— C
ON
FIR
Mur
sig
s 55
9 IN
FO
ana
me
hr d
afyd
d IN
FO
bqt
h nr
lon
don
INF
Oc
hw?
CO
NF
IRM
w4b
c de
g4f
hz
IDE
NT
kn
CL
OS
UR
Eg4
fhz
de W
4bc
IDE
NT
r es
gm
daf
ydd
CO
NF
IRM
ur r
st 5
49 q
rm
INF
Oa
hand
le h
r tr
acy
INF
Ob
qth
fla
INF
Oc
hw?
CO
NF
IRM
g4fh
z de
\v4
bc
IDE
NT
kn
CL
OS
UR
E
W4b
c de
g4f
hz
IDE
NT
r tr
acy
fb c
py
CO
NF
IRM
qsl
? gl
es
73
CO
NF
IRM
W4b
c d
e g4
fhz
IDE
NT
sk
CL
OS
UR
E
g4fh
z de
W4b
c ID
EN
Tr
da
fyd
d C
ON
FIR
Mw
l qs
l hp
e cu
agn
—
CO
NF
IRM
g4fh
z de
W4b
c ID
EN
Tsk
CLOSURE
-IN
FO
-
EN
D-
CO
NT
R-•
(CONTR-
>CONTR-•
\CONTR-
^EXCHANGE-
\EX
CH
AN
GE
\EX
CH
AN
GE
^U
TT
ER
AN
CE
. C
ON
TE
XT
>
"7T
RA
NS
AC
TIO
N-
a a 03 S3 o z
-UP
TA
KE
- T
OP
IC-
-|i
PE
RS
ON
|-
-[2
PE
RS
ON
| -
-CH
AN
NE
L —
- P
ER
SO
N-
Fig
ure
5: S
truc
ture
ass
igne
d by
con
stit
utiv
e sy
stem
to
a co
nstr
ucte
d m
inim
al I
AR
MT
tra
nsac
tion
.
IDIOMATICITY AND FUNCTIONAL VARIATION
nity of PERSONS, and the communication breakdown would be marked byexcessive use of uptake loops.
4. CONCLUSION: IART AND OTHER VARIETIES
The present approach synthesizes an account of the linguistic problem of idioma-ticity with selected aspects of dialogue analysis in order to account for functionallanguage variation on a coherent functionalistic basis. Cover terms such as "reg-ister" (section 1.2) are replaced by more fundamental concepts, and variationthat involves adaptation to and modification of contexts of utterance is analyzedusing the "restrictive register" of IART. This approach is related to a largenumber of other parametric accounts of language variation (cf. Crystal & Davy1969; Ervin-Tripp 1964; Hymes 1972; Klein 1974); it differs from themmainly in its explicit functionalistic foundations and in the intensive treatment ofa single register. A specific feature of the approach is that each of the propertiesand relations assigned to constitutive categories (section 1.2) is a value of acommunication parameter (e.g., face-to-face vs. teleglossic); each parameter isscalar and weighted or "marked" toward one end (here the teleglossic end ismore "marked") in terms of the restrictions it imposes on utterances. Theregions within the "quality space" defined by these parameters constitute spe-cific registers, some of which are more "restrictive," some less "restrictive,"depending on the total "markedness" of the parameter values involved.
A more restrictive neighbor of IART in this quality space is Morse telegraphy;similar to IART are official two-way radio services, less restrictive is CB (Smith1979) or the telephone. Less restrictive in SETTING (non-null) but more inCHANNEL use (no turn-taking) are broadcast radio and television from the pointof view of listener or viewer. Other registers resemble IART less, with fewerrestrictive properties and less stereotypy. Kinds of language use with pronouncedsimilarities to IART along one or more dimensions include writing, ritual orceremonial speech (Arewa & Dundes 1964, esp. 8off.; Sacks et al. 1974: 710),classroom teaching (Sinclair & Coulthard 1975), advertising (Leech 1966),nonce communication between strangers with little if any knowledge of eachother's language (high "user noise"), pidgins (Hymes 1971), and baby talk(Ferguson 1977). These suggestions cannot be documented in detail here, but byway of illustration it may be noted that, in written correspondence, turn-takingpatterns and stereotypy in framing resemble properties of IART in being relatableto the unimodal, simplex, technically aided teleglossic CHANNEL (section 2.3).Differences in propagation speed (e.g., affecting uptake loop occurrence) resultfrom visual storage properties of the written medium.
It is profitable to speculate on possible extensions of the present approach inother areas of language use and variation. For instance, the possibility of ac-counting for some aspects of language acquisition in terms of progression fromstereotypic (as automatic, imitative, routine, gestalt, holistic) "restrictive"
39
DAFYDD GIBBON
speech to mixed restrictive and creative speech is interesting (Krashen & Scar-cella 1978; Peters 1977; Fillmore 1976; also the system expansion modeled byHalliday 1975). The functional cycle (cf. section 1.2) is also related to the kindsof monitoring that are postulated in sociopsychological studies such as those ofLabov (1972), in which parallels between natural and conventional contextualconstraints were held to relate to just such a function.
ACKNOWLEDGMENTS
For discussion of points from an earlier version of this paper ('"High-level idiom: Analysis of arestrictive register," 1978) I am indebted to the editor of this journal, to two anonymous readers, andto my students and colleagues at the universities of Bielefeld and Gottingen, particularly ChristianeDechert, Margret Selting, and Gerhard von der Straten; to Rosemarie Tracy for helping to clarifymany obscurities "mni tnx es 88 hi ," and to those who taught me 1ART "tu oms fr info es qsl dedj0mz/g4fhz."
REFERENCESArewa, E. O , & Dundes, A. (1964). Proverbs and the ethnography of speaking folklore. In J. J.
Gumperz & D. Hymes (eds.). The ethnography of communication. American AnthropologistSpecial Publication 66:6, Pt. 2. Menasha: American Anthropological Association, 70-85.
A.R.R.L. (1977). The radio amateur's handbook. Newington, Conn.: American Radio RelayLeague.
Austin, J. L. (1962). How to do things with words. London: Oxford University Press.Blom, J.-P., & Gumperz, J. J. (1972). Social meaning in linguistic structures: Code-switching in
Norway. In J. J. Gumperz & D. Hymes (eds.). Directions in Sociolinguistics. N. Y.: Holt, Rine-hart & Winston, 407-34.
Bolinger, D. L. (1950). Rime, assonance and morpheme analysis. Word 6: 117-36.Bublitz, W. (1978). Ausdrucksweisen der Sprechereinstellung im Deutschen und Englischen. Unter-
suchungen zur Syntax, Semantik und Pragmatik der deutschen Modalpartikeln und Vergewis-serungsfragen und ihrer Englischen Entsprechungen. Tubingen: Niemeyer.
Biihler, K. (1934). Sprachtheorie. Die Darstellungsfunktion der Sprache. Stuttgart: Fischer.Burger, H. (1973). Idiomatik des Deutschen. Tubingen: Niemeyer.Chao, Y.-R. (1956). Tone, intonation, singsong, chanting, recitative, tonal composition and atonal
composition in Chinese. In M. Halle, H. Lunt, & H. McLean, (eds.), For Roman Jakobson. TheHague: Mouton. 52-59.
Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, Mass.: M.l.T. Press.Crystal, D., & Davy, D. (1969). Investigating English style. London: Longman.DaneS, F. (i960). Sentence intonation from a functional point of view. Word 16: 34-54.Duncan, S. (1973). Toward a grammar for dyadic conversation. Semiotica 9: 29-46.Ellis, J., & Ure, J. N. (1969). Language varieties: Register. In A. R. Meetham, ed.. Encyclopaedia
of Linguistics, Information and Control. London: Pergamon Press. 251-9.Ervin-Tripp, S. (1964). An analysis of the interaction of language, topic, and listener. In J. J.
Gumperz & D. Hymes. (eds.). The ethnography of communication. American AnthropologistSpecial Publication 66:6, Pt. 2. Menasha: American Anthropological Association, 86-102.
Ferguson, C. A. (1977). Baby talk as a simplified register. In C. E. Snow & C. A. Ferguson, (eds.).Talking to Children. Language Input and Acquisition. Cambridge: Cambridge University Press,209-35.
Fillmore, L. W. (1976). Individual differences in second language acquisition. Asilomar Conferenceon Individual Differences in Language Ability and Language Behavior. Monterey. Calif.
Firth, J. R. (1968). Linguistics and translation. In F. R. Palmer, (ed.), Selected Papers of]. R. Firth1952-59. London: Longman, 84-95.
Fraser, B. (1970). Idioms within a transformational grammar. Foundations of Language 6: 22-42.
40
IDIOMATICITY AND FUNCTIONAL VARIATION
Gibbon, D. (1976a). Perspectives of intonation analysis. Bern: Lang.(1976b). Performatory categories in contrastive intonation analysis. In D. Chitoran (ed.).
Second International Conference of English Contrastive Projects. Bucharest: Bucharest UniversityPress and Washington: Center for Applied Linguistics, 145-56.
_. (in press). Violations of Frege's Principle and their significance for contrastive seman-tics. Papers and Studies in Contrastive Linguistics (Poznaii).
Giilich, E., & Henke, K. (1980). Sprachliche Routine in der Alltagskommunikation. Uberlegungenzu "pragmatischen Idiomen" am Beispiel des Englischen und Franzosischen. Vol. I, II. DieNeueren Sprachen 78 (1979): 513-30; 79 (1980): 2-33.
Hall, E. T. (1964). Adumbration as a feature of intercultural communication. In J. J. Gumperz & D.Hymes (eds.), The ethnography of communication. American Anthropologist Special Publication66:6, Pt. 2. Menasha: American Anthropological Association, 137-53.
Halliday, M. A. K. (1967). Notes on transitivity and theme in English. Journal of Linguistics 2:37-81; 3: 199-244; 4: 179-215-
(1975)- Learning how to mean: Explorations in the development of language. London:Arnold.
Halliday, M. A. K., Mclntosh, A., & Strevens, P. (1964). The linguistic sciences and languageteaching. London: Longman.
H.M.S.O. (1938). General Post Office handbook for wireless operators working installationslicensed by His Majesty's Postmaster General. London: His Majesty's Stationery Office.
Hockett, C. F. (1955). A manual of phonology. International Journal of Linguistics, Memoir 11.Baltimore: Waverley Press.
(1956). Idiom formation. In M. Halle, H. Lunt, & H. McLean, (eds), For Roman Jakob-son. The Hague: Mouton, 222-9.
_. (1958). A course in modern linguistics. New York: Macmillan.Hymes, D., (ed.). (1971). Pidginization and creolization of languages. Cambridge: Cambridge
University Press.(1972). Models of the interaction of language and social life. In J. J. Gumperz, & D.
Hymes, (eds.), Directions in sociolinguistics: The ethnography of communication. New York:Holt, Rinehart & Winston, 35-71.
-• ('975)' Pre-war Prague school and post-war American anthropological linguistics. In E. F.K. Koerner, (ed.), The transformational-generative paradigm and modern linguistic theory.Amsterdam: John Benjamins, 359-80.
Jakobson, R. (i960). Closing statement: Linguistics and poetics. In T. A. Sebeok, ed., Style inlanguage. Cambridge, Mass.: M.I.T. Press, 350-77.
Katz, J. J., & Postal, P. M. (1963). Semantic interpretation of idioms and sentences containing them.M.I.T. Research Laboratory of Electronics Quarterly Progress Report 70: 275-82.
Klein, W. (1974). Variation in der Sprache. Ein Verfahren zu ihrer Beschreibung. Kronberg:Scriptor.
Krashen, S., & Scarcella, R. (1978). On routines and patterns in language acquisition and perfor-mance. Language Learning 28.
Labov. W. (1972). Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.Ladd, D. L. (1978). Stylized intonation. Language 54: 517-40.Leech, G. N. (1966). English in advertising: A linguistic study of advertising in Great Britain.
London: Longman.Lewis, D. (1972). General semantics. In D. Davidson & G. Harman, (eds.), Semantics of natural
language. Dordrecht: Reidel, 129-218.Makkai, A. (1972). Idiom structure in English. The Hague: Mouton.Malkiel, Y. (1959). Studies in irreversible binomials. Lingua 8: 113-60.Marchand, H. (i960). Categories and types of English word-formation: A synchronic-diachronic
approach. Wiesbaden: Harrassowitz.Newmeyer, F. J. (1974). The regularity of idiom behavior. Lingua 34: 327-42.Peters, A. M. (1977). Language learning strategies: Does the whole equal the sum of the parts?
Language 53.Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking for conversation. Language 50: 696-735.Sadock, J. M. (1974). Toward a linguistic theory of speech acts. New York: Academic Press.
41
DAFYDD GIBBON
Searle, J. R. (1969). Speech acts: An essay in the philosophy of language. Cambridge: CambridgeUniversity Press.
(1976). Review of Sadock (1974). Language 52: 966-71.Shannon, C. E. (1949). The mathematical theory of communication. In C. E. Shannon & W.
Weaver, (eds.), The Mathematical Theory of Communication. Urbana: University of IllinoisPress, 29-125.
Sinclair, J. M., & Coulthard, R. M. (1975). Towards an analysis of discourse. The English used byteachers and pupils. London: Oxford University Press.
Smith, J. J. (1979). Male and female ways of speaking: Elaborately restricted codes in a CB speechcommunity. Papers in Linguistics. International Journal of Human Communication 12: 163-84.
Smith, L. P. (1925). Words and idioms: Studies in the English language. London: Constable.Weinreich, U. (1966). Problems in the analysis of idioms. In J. Puhvel, ed., Substance and structure
of language: Lectures delivered before the Linguistic Society of America at UCLA ig66. Berkeley:University of California Press, 114-71.
Yngve, V. (1970). On getting a word in edgewise. Papers from the Sixth Regional Meeting of theChicago Linguistic Society, 567-78.
42
